Rate limits

Every workspace has three separate limits: monthly call budget, requests-per-second, and concurrent requests. Each ticks independently.

Limits by plan

Response headers

Every response includes usage state:

X-RateLimit-Limit: 100
X-RateLimit-Remaining: 87
X-RateLimit-Reset: 1712339400
X-Concurrent-Limit: 50
X-Concurrent-Active: 12

When you hit a limit

Exceeded requests return HTTP 429 with a Retry-After header (seconds). We recommend exponential backoff starting at the suggested retry value.

HTTP/1.1 429 Too Many Requests
Retry-After: 3
Content-Type: application/json

{
  "error": "rate_limited",
  "message": "100 req/s ceiling hit; retry in 3s",
  "retry_after_ms": 3000
}

Concurrency vs. rate

Rate limits count request starts. Concurrency counts in-flight browsers. A long crawl consumes concurrency for its whole duration but only one rate slot at launch.